Lucas Mechanical PI System Model

[Roger B]

The pieces of the test fixture were drilled and tapped as appropriate. This fixture was mounted on the column of the gear pump test fixture and I made a second injector pipe so the volumes could be measured.

The pump was assembled with all the O rings and seals, the stop plug avoided damage to the sleeve O ring.

The shuttle and adjustable stop were then fitted along with the adjuster end cap. With the drive end cover and Oldham coupling fitted the pump turned smoothly and with the motor and brackets in place it would rotate down to quite low speeds without problems 

<a href="https://www.youtube.com/watch?v=dkLo2QjDPF0" target="_blank">http://www.youtube.com/watch?v=dkLo2QjDPF0</a>

[Roger B]

The complete system was then assembled with variable power supplies for the gear pump and shuttle pump.

I started without the delivery valves in place to bleed the air out of the system. The pulses from the unions were visible.

<a href="https://www.youtube.com/watch?v=bNZ4CLM3Uws" target="_blank">http://www.youtube.com/watch?v=bNZ4CLM3Uws</a>

I then fitted the delivery valves and injector pipes so I could collect the fuel in a pair of measuring cylinders. There was a lot of leakage from the pump, probably due to the fit (or lack of) between the shuttle rotor and sleeve  . These are just silver steel in reamed holes for this first attempt. On the positive side the pump did what it should and the volumes delivered were equal from both outlets 

https://youtube.com/shorts/Kc2S5TK5cbo

<a href="https://www.youtube.com/watch?v=Kc2S5TK5cbo" target="_blank">http://www.youtube.com/watch?v=Kc2S5TK5cbo</a>

[cnr6400]

 

[Vixen]

Hello Roger,

Excellent test result from the first run.   

I cannot wait for you to add the two low pressure injectors and throttle up and down on the shuttle. Are you going to need to reduce the body leaks before you attempt that or just go ahead and sort that problem out later

Cheers
Mike

[Admiral_dk]

Nice to see that the Metering gives to Equal amounts of Fuel as predicted 
That bodes well a Linear Metering between the Adjustable Stroke and Fuel too 

As Mike, I hope that adding the Injectors keeps the same relationship between parts 

Per

[Roger B]

Thank you all for your support 

I will need to reduce the fuel leakage before the next tests. The increased pressure with the injectors fitted will only increase the problem.

First, I needed to investigate the leakage. I removed the rotor from the pump and replaced it with the length of 6mm silver steel that the rotor was made from. This eliminated any leakage via the shuttle components. Turning on the feed pump showed rapid leakage between the rotor and the sleeve:

<a href="https://www.youtube.com/watch?v=qAP36M4eUKc" target="_blank">http://www.youtube.com/watch?v=qAP36M4eUKc</a>

This was not unexpected as the Tolerance on the reamed hole and the silver steel are quite large (in this context). Measuring what I had was not easy, but the silver steel was close to 0.01mm under size and the bore probably 0.01mm oversize (OK for H7). This gave a total clearance of maybe 0.02mm which for a 6mm diameter calculate to equivalent to a 0.1mm diameter hole. Not unexpectedly I need to move to hardened and lapped components. I already have the tools and experience to produce the 2 mm shuttle components (same as my other fuel injection pumps). The 6 mm parts would need some new tools and some learning.

My experience with the smaller components suggests that the silver steel grows slightly when hardened so I made a new rotor, as before, and hardened it. It appeared slightly bigger but that could have been the scale film. I hardened it in a gas flame, quenching it after it passed through the ‘hold point’. I didn’t try and soak it for 15 mins as the Tubal Caine book suggests to achieve maximum hardness.

I then made a simple external lap from 16mm brass rod. The bore was carefully reamed 6mm. The original rotor would pass through, the hardened one would not. The lap was then finished with an expanding grubscrew and a 1mm slit.

[Roger B]

To hold the rotor for lapping I tapped an M4 x 0.5mm thread in the end of a length of 5mm rod. I then lapped it with 5 micron diamond paste until the surface was smooth and clean.

Measuring is not so easy with my simple equipment. The stock bar remains under 6 mm and the hardened rotor is definitely over 6 mm. The hardened rotor just enters the original sleeve whereas the original feels quite slack.

I now need to look at suitable pin gauges, maybe under size reamers and a 6mm Acrolap to make a better sleeve bore.

[Admiral_dk]

I suppose that there isn't any other way - like using seals - as the friction will rise rather quickly (or not ? ) ....

You already know more than most of us, about how to make very fine tolerance parts - so I can't attribute anything, but my best wishes for success 

Per         

[Laurentic]

Just idly thinking - probably rubbish, but thinking out the box - but if a new rotor was made using the tad under 6mm stock, and that then had its diameter reduced a tad by whatever means, and then that was hardened, would it come out at spot-on the desired fit?

Chris. 

[petertha]

Roger, I highly recommend an indicating micrometer. It makes quick work of comparing dimensions, especially relative to a target standard. Here, analog needle wins hands down to digital. You can get some great deals on used ones relative to what they cost new. I'm kind of a Mitutoyo fan boy so with a bit of patience found a nice used 0-1" (tenths reading) for ~$50 CAD & a bit more for 1-2" Mitutoyo in 'storage shelf' condition. One was still encased in the original rubbery protection dip (brand new). I noticed, at least here, for every one imperial indicating mic there seem to be 3 metric ones available for some reason. And metric offerred many other quality brand names, more to choose from. They have same / higher graduation resolution.

My experience with drill rod, at least here in N-AM is they are within the stated tolerance +/- 0.0005" IIR. But they are inevitably eliptical shape which can account for some of this deviation. Lapping it round isnt quite as straightforward unless the process can be guaranteed to remove the hill tops (only). Some of my simple home shop clamshell type lapping tools were not doing this as effectively as I'd hoped (only apparent with an indicating mic).

Following along your nice work, keep it up!

[Vixen]

My experience with drill rod, at least here in N-AM is they are within the stated tolerance +/- 0.0005" IIR. But they are inevitably eliptical shape which can account for some of this deviation. Lapping it round isnt quite as straightforward unless the process can be guaranteed to remove the hill tops (only). Some of my simple home shop clamshell type lapping tools were not doing this as effectively as I'd hoped (only apparent with an indicating mic).

Roger, I agree with petertha. Centreless ground rod is usually trilobed as opposed to elliptical in cross section; definitely NOT round. There are usually three hill tops. Amazingly the diameter is always within the stated tolerance, when measured with a standard two point of contact micrometer. But the ground bar is NOT round, it is a trilobed (trianglular??) with three hill tops. Unfortunately a simple two point of contact external lapping toll will NOT pull it back to round. You will need a more sophisticated round lapping tool to remove those hill tops.

Mike

[Roger B]

Thank you all for your thoughts and suggestions 

I am aware of the trilobar problem with centreless grinding as I had a bad piece of 2mm silver steel during my diesel work. That is why I moved to using pieces of pin gauges for my injection pump plungers.  I am hoping that my close-fitting round laps will help to remove the peaks, But the root diameter will tend to end up too small.

I was interested to see what had happened to the bore on the hardened rotor. A quick check between my pin gauges and the micrometer showed that things remained consistent. The 2.00 mm pin gauge passed easily through the bore on the original rotor that had been reamed 2 mm.

The 1.95 mm gauge would not enter the bore of the hardened rotor, reamed 1.98 mm. I then started to lap this with a 70 thou Acrolap and 5 micron diamond paste. After a short time lapping the 1.95 mm pin gauge would enter from both ends, but would not pass more than ¾ of the way through. This usually means that the bore is not straight   

I was not able to properly check the 2mm silver steel as I did not have a small enough V block, it just fell into the grove at the base of the V. I could, however, check the 6 mm silver steel. I made a rough measuring set up using my angle plate, the magnetic dial gauge stand doesn’t stick to my granite surface plate   

The 6 mm stock was straight and round:

<a href="https://www.youtube.com/watch?v=5Xvb16eE8b8" target="_blank">http://www.youtube.com/watch?v=5Xvb16eE8b8</a>

The hardened rotor was not good   
 
<a href="https://www.youtube.com/watch?v=GeBH8HNmurU" target="_blank">http://www.youtube.com/watch?v=GeBH8HNmurU</a>

So which way to go from here?

I thought about using the stock materials and sealing the rotor with O rings. This has some problems:

Will the required 4 x 1 mm O rings be robust enough?

I will have to seal both sides of the ports otherwise the leakage between the ports, which will be at full pressure, may interfere with the metering at low volumes.

How can I assemble it without slicing pieces off the O rings when the slide past the ports?


I think I will have to use unhardened parts, lapped to a close fit. The 6 mm silver steel can be nearly 0.01 mm under size, after lapping it will be less. I would then need to be able to ream the sleeve bore to a suitable size so I can lap it to fit the rotor. I can buy reamers in 0.01 mm steps, but they are not cheap, so I would like to be right first time.

[Vixen]

Hello Roger,

The only leak proof, lapped shaft/body rotatable 'seal' I know of, is tapered. You find them on steam cocks, drain cocks and in glass laboratory equipment taps. They can be lapped to be completely leak proof, however they can also lock, which prevents rotation with only a small pressure applied to the wide end.

A simple parallel lapped rotary 'seal' between shaft and body, by the very nature of being free to rotate, must have some clearance, however small. You are always going to get some leakage through a simple parallel shaft 'seal'. You will be leaking fuel, so even a small leak is not a good idea.

Some form of rotary lip seal(s) between the body and the rotating shaft should be able to eliminate the shaft/ body fuel leak. You could also look a spring loaded carbon/ ceramic face seals.

I have also heard of similar rotary shaft 'seals' having grooves cut into the shaft, which are pumped with pressurised oil. The higher oil viscosity in the grooves, seal the shaft from leaking fuel. The oil pressure needs to be the same or slightly higher than the fuel pressure, it's then a balancing act between a fuel leak and an oil leak. 


I fully understand the additional size of the seals and added manufacturing complications are undesirable. Sounds like you find yourself between a rock and a hard place yet again, but that's model engineering

Mike

[Minh Thanh]

I decided to make the new injectors with a 1.2mm basis. It is slightly easer than 1mm and the needles can still be made from 1.5mm silver steel. 1mm may be relevant for the diesel systems but I don’t see any difference here.

The two nozzles and the needle were made as before. I just made one needle in case the angle needs to be widened. I adjusted the spring pressure to gest an opening pressure of around 4 bar which seemed to work quite well. There was none of the locking I had seen with the narrow angle needles. The needle has not been lapped to the seating in this video clip. It may not be necessary for petrol 

https://youtube.com/shorts/tbQXu2DdWgk?feature=share

<a href="https://www.youtube.com/watch?v=tbQXu2DdWgk" target="_blank">http://www.youtube.com/watch?v=tbQXu2DdWgk</a>

 I understand the difficulty with making small injectors and pump systems.

Because you're using a gear pump system – which is different from a system using a plunger – if it were me, I would make a pressure valve. That valve only opens when the fuel pressure is high enough, e.g., 4 bar, 5 bar, etc., which creates instantaneous pressure for the injector, facilitating atomization. With small injectors, everything becomes difficult to control.

[Minh Thanh]

Hi Roger
To be clear,

in your video: the injector seals well at low pressure. When the pressure increases, fuel starts to leak and oil droplets appear. This happens several times, and the injector only sprays when the pressure reaches its highest point after the leak phase - meaning the pressure is high enough for the injector to open fully -  the leak when the fuel pressure reaches a point close to the spring pressure.
Increasing the spring force might solve the problem, but with these small springs, it's difficult to control it to open precisely within a pressure range, for example, 4.5 --> 4.8 bar - under conditions where the pressure generated by the pump increases linearly.
Personally, I would try to create a high and almost instantaneous pressure. Adding a pressure valve, for example, around 6 bar, and adjusting the injector spring to around 3 bar might solve the problem - but I'm not sure because I haven't tried it.